TWI420417B - Method of and apparatus for processing image data for display by a display device - Google Patents

Description

Method and apparatus for processing image data displayed by a display device

The present invention relates to a method and apparatus for processing image data displayed by a display device.

Active matrix liquid crystal display devices are known that can switch between a common display mode and a private display mode, or can direct multiple different images to different viewers. In the first (common) mode, these displays typically behave as standard displays. The device displays a single image to the widest possible range of viewing angles for optimal brightness, image contrast and resolution for all viewers. In the second (private) mode, the main image can be discerned only from a reduced range of viewing angles (typically centered on the normal to the display surface). From this reduced angle range, the viewer of the external gaze display will perceive the second occlusion image of the main image, or be downgraded to cause an incomprehensible main image.

In GB2413394 (Sharp), a switchable privacy device is constructed by adding one or more additional liquid crystal layers and polarizers to the display panel. The inherent viewing angle dependence of such additional elements can be altered by electrically switching the liquid crystals in a known manner. Devices utilizing this technology include the Sharp Sh851i and Sh902i mobile phones.

However, these types of displays can only selectively attenuate the brightness of the main image for observers outside of the reduced viewing range and cannot display reconfigurable color video images to the side viewer. A multiview display based on parallax barrier technology that does have this capability is described in US 2007/0296874 A1 and US Pat.

All of the above methods suffer from the disadvantage that they require additional equipment to be added to the display to provide functionality to electrically switch the range of viewing angles. This adds cost to the display and in particular adds volume to it, which is very poor (especially in mobile display applications such as mobile phones and laptops). The display based on the parallax barrier can display only half of the pixels in the display to each viewing area even in the common mode, and therefore, the effective image resolution is half of the resolution of the base panel.

An example of a display device having privacy mode capabilities without increased display hardware complexity is the Sharp Sh702iS mobile phone. This is combined with the angular data illumination property inherent in the liquid crystal mode in the display to manipulate the image data displayed on the LCD of the telephone to produce a private display that is not understandable by the viewer of the self-eccentric position observation display. mode. However, the quality of the image displayed to the legal on-axis viewer in the private mode is somewhat degraded.

A multiview display based solely on image processing technology is described in GB2428152A1 and in GB Patent Application No. 0804022.2, which is capable of providing detailed reconfigurable side images in a private mode without the need for a display panel. Extra optical equipment. In such displays, the primary image data is manipulated in a manner that relies on the second occlusion image, and thus causes the off-axis viewer to perceive the occlusion image when the modified image data is displayed on the panel.

As described in GB Patent Application No. 0804022.2, high spatial frequency modulation applied to the main image to produce a multi-view effect can result in undesirable color artifacts that are apparent in the main image. The above application describes the application of an image processing filter (such as the image processing filter described in GB Patent Application No. 0701325.3 (published as GB-A-2445982)) to the main image so as to blur the fine features in the image. The appearance of these artifacts is generally improved. However, standard image blur filters (including the standard image blur filters described for multi-view displays (such as the above disclosure) are not designed to specifically address the particular color associated with the image data manipulation multiview process. False shadow. As a result, it has been shown to be insufficiently effective in removing color artifacts with minimal blurring of the main image for all image types. In particular, the filter cannot correct color artifacts around black text and white text by the same set of parameters.

Accordingly, it is desirable to provide a multi-view display that substantially unaltered the common mode relative to an equivalent standard display panel (ie, full brightness, resolution, contrast, viewing range, etc.) and that has an image-based processing method Multi-view or privacy mode capability, wherein the quality of the main image as observed by the main viewer is maximized by including image processing steps for specifically correcting the type of color artifacts produced by the multi-view image processing program. Jiahua.

According to a first aspect of the present invention, a method for processing image data displayed on a display panel of a display device is provided, comprising: receiving image pixel data representing an image; and processing pixel data in a first processing step The viewer produces a multi-view effect; and in a second processing step, each of the plurality of subsets of pixel data, wherein each subset includes the same number of pixel groups and each pixel group includes at least one pixel, Deriving new pixel data for at least one of the subset of pixels of the subset based on the pattern of pixel data in the subset of pixels of the subset, the new pixel data being derived to distinguish at least one of the patterns from the reverse pattern thereof The way it is implemented.

The first processing step can be performed in consideration of the nature of the display panel in order to introduce the following changes: changes in the on-axis illuminance, which tend to be locally balanced to the on-axis viewer via spatial averaging and thus will not be viewed on-axis It is perceived by the person; and the change in the out-of-axis illumination, which is not locally balanced by the spatial averaging to the off-axis viewer and will therefore be perceived by the off-axis viewer.

According to a second aspect of the present invention, a method for processing image data displayed on a display panel of a display device is provided, comprising: receiving image pixel data representing an image; and considering a property of the display panel in the first processing step Pixel data is processed to introduce the following changes: changes in illumination on the axis that tend to be locally balanced to the on-axis viewer via spatial averaging and therefore will not be perceived by the on-axis viewer; and off-axis illumination a change that is not locally balanced by the off-axis viewer via spatial averaging and thus will be perceived by the off-axis viewer; and in the second processing step, for each of the plurality of subsets of pixel data Each of the subsets includes the same number of pixel groups and each pixel group includes at least one pixel, and at least one of the subset of pixel groups is dependent on the pattern of pixel data in the pixel group of the subset One to export new pixel data.

In the first processing step, any illuminance increase introduced by the viewer for one of a pair of pixel groups having a single illuminance via spatial averaging may be by the pixel group for the pair The substantially equivalent illuminance reduction of the other of the groups is substantially matched.

It can be configured as follows: the obtained illuminance of one of the pair of pixel groups is close to the maximum illuminance, or the obtained illuminance of the other pixel group of the pair is close to the minimum illuminance.

The deriving the new pixel data can include selecting a filter dependent on the pattern of pixel data, and applying the selected filter to at least some of the pixel groups in the subset of pixels to derive new pixel data.

The method can include comparing the pattern of pixel data with a plurality of predetermined patterns, and deriving new pixel data depending on the result of the comparing step.

Each of the predetermined patterns can be associated with a respective respective filter, and the method can include determining a matching pattern in the comparing step and selecting its associated filter for deriving new pixel data.

The comparing step may include: determining, for each of the at least one pixel group or the at least one pixel group for which the new pixel data is derived, determining the brightness of the pixel group relative to the at least one immediately adjacent pixel group The measure, which distinguishes between higher brightness and lower brightness.

The method can include: assigning a group of pixels to one of a predetermined set of levels according to pixel data of the group of pixels for each group of pixels in the subset; and comparing the pattern of the assigned levels with the comparing step Schedule a pattern.

The method can include calculating a metric based on pixel data of at least some of the pixel groups in the subset of pixels, and can perform the step of assigning the group of pixels to one of the predetermined levels, depending on the calculated metric.

According to a third aspect of the present invention, a method for processing image data displayed on a display panel of a display device, comprising: receiving image pixel data representing an image; and processing pixel data in a first processing step The viewer produces a multi-view effect; and in a second processing step, for each of the plurality of subsets of pixel data, each subset includes the same number of pixel groups and each pixel group includes at least one pixel: Metrics are calculated based on pixel data of at least some of the pixel groups in the subset of pixels; for each pixel group in the subset, pixel groups are assigned based on pixel data of the pixel group and relying on calculated metrics Up to one of a predetermined set of levels; comparing the pattern of the assigned levels with a plurality of predetermined patterns; and deriving new pixel data for at least one of the subset of pixels of the subset depending on the result of the comparing step.

This metric can be an average.

The method can include assigning a group of pixels to one of a predetermined set of levels based on a comparison between pixel data of the group of pixels and the calculated metric (eg, using the calculated metric as a threshold).

The method can include, for example, converting the pixel data to a illuminance value based on a gamma power law function for use in the deriving step.

The method can include using at least one conversion parameter determined by a pattern of pixel groups that depend on the subset.

The subset of pixels of the subset may be contiguous and extend substantially in one dimension.

A subset of pixels can include a connected two-dimensional configuration.

Each of the plurality of subsets may be spaced apart from another subset by a group of pixels.

The step of exporting the new pixel data may use the knowledge of the processing performed in the first processing step.

New pixel data may be derived in a manner that accounts for the pattern of pixel data representing the image features that tend to cause artifacts due to processing performed in the first processing step.

The deriving step may include determining whether each of the at least one pixel group or the at least one pixel group for which the new pixel data is derived forms a part of a pattern of pixel data representing the image feature, the image feature being in the first process The processing performed in the steps tends to result in artifacts.

The pattern of pixel data representing the bright features on the relatively dark background may be processed in the export step in a manner different from the pattern of processing the pixel data representing the dark features on the relatively bright background.

The deriving step may include, from the pattern determination of the pixel data, deriving whether each of the at least one pixel group or the at least one pixel group for which the new pixel data is formed forms part of at least one of the following image features: having a single a dark or bright line of the width of a group of pixels; a group of pixels adjacent to a dark or bright line having a width of a single group of pixels; a left or edge of a dark or bright line having a width of two groups of pixels; Or a dark-light checkerboard pattern between two pixel groups; and a diagonal line.

The second processing step can be performed prior to the first processing step.

The second processing step can be performed after the first processing step.

New pixel data can be derived for a single group of pixels in a subset of pixels of the subset.

Each pixel group may comprise a composite color pixel group of color component pixels, the method being applied sequentially to each of the color component pixels.

The composite color pixel group can include red, green, and blue component pixels.

A first processing step can be performed to interleave a set of received pixel data representing different individual images to present different images to viewers in different respective viewing locations.

According to a fourth aspect of the present invention, there is provided an apparatus configured to perform the method according to any one of the first to third aspects of the present invention.

According to a fifth aspect of the invention, there is provided a display device comprising a device according to a fourth aspect of the invention.

According to a sixth aspect of the present invention, there is provided a program for controlling a device to perform the method according to any one of the first to third aspects of the present invention. The program can be carried on the carrier medium. The carrier medium can be a storage medium or a transmission medium.

The nature of the display panel under consideration may be the response of its signal voltage to on-axis illuminance, wherein the illuminance variation introduced on the axis is configured to tend to be locally balanced to the on-axis viewer via spatial averaging and thus Will not be noticed by the on-axis viewer. The panel will have a non-linear out-of-illuminance versus on-axis illuminance such that the changes introduced in the off-axis illumination will not be locally balanced to the off-axis viewer via spatial averaging and will therefore be an off-axis viewer Perceived.

An embodiment of the present invention provides an improved image processing method for a private and multi-view display that allows such displays to be introduced without the image artifacts that are not desired by the multi-view data manipulation processing program. Maintaining high-definition image quality, these displays are types that use image data manipulation and the inherent angular viewing dependence of the LCD panel to allow multiple images to be viewed by multiple viewers.

In a first embodiment of the present invention, at least one of the plurality of image data sets input to the device is processed by a filter to detect a specific image that is known to cause image artifacts when performing the multi-view image processing program. feature. The particular image feature is then manipulated in a manner that prevents the introduction of artifacts while maintaining as much detail as possible of the original image, depending on the type of particular image feature.

In a second embodiment of the present invention, a group of pixels of a localized pixel group in an output image caused by a multiview image processing program and a same region corresponding to at least one of the input images is compared. The difference is then detected and the multi-view output image is altered in a manner that calibrates the difference or spreads it over a wider area of the multi-view output image to make it less noticeable.

In a preferred embodiment, the display is comprised of a standard LCD display with modified control electronics. LCD displays typically consist of several component parts, including:

1. A backlight unit for supplying uniform wide angle illumination to a panel.

2. A control electronics for receiving digital image data and outputting an analog signal voltage for each pixel, and a timing pulse and a common voltage for the opposite electrodes of all of the pixels. A schematic of a standard layout of LCD control electronics is shown in Figure 1 (Ernst Lueder, Liquid Crystal Displays, Wiley and sons Ltd, 2001).

3. An LC panel for displaying an image by spatial light modulation, consisting of two opposite glass substrates, a pixel electrode array and an active matrix array disposed on one of the glass substrates to control the electronic device The received electronic signal is directed to the pixel electrode. A uniform common electrode and a color filter array film are typically disposed on another substrate. A liquid crystal layer of a given thickness (typically 2-6 μm) is contained between the glass substrates, which can be aligned by the presence of an alignment layer on the inner surface of the glass substrate. The glass substrate will typically be placed between the crossed polarizing film and the other optical compensation film to cause an electrically induced alignment change in each pixel region of the LC layer to produce the desired optical modulation of light from the backlight unit and the surrounding environment. Change and thereby generate an image.

An embodiment of the invention disclosed in GB Patent Application No. 0804022.2, which operates in a common display mode, is schematically illustrated in FIG. In general, LCD control electronics 1 (also referred to herein as control electronics) will be specifically configured for the electro-optic characteristics of LC panel 2 to be viewed for a master view from a direction perpendicular to the display surface (on-axis) The output voltage dependent on the input image data is output in a manner that optimizes the perceived quality of the displayed image (ie, resolution, contrast, brightness, response time, etc.). The relationship between the input image data value of a given pixel and the observed illuminance caused by the display (gamma curve) is the combined effect of the mapping of the data value of the display driver to the signal voltage and the response of the signal voltage of the LC panel. Make a decision.

The LC panel 2 will typically be configured with multiple LC domains per pixel and/or a passive optical compensation film to keep the display gamma curve as close as possible to the on-axis response for all viewing angles, thereby providing a wide viewing area 5 Provides substantially the same high quality image. However, the inherent property of liquid crystal displays is that their electro-optic response is angularly dependent and the off-axis gamma curve will be different from the on-axis gamma curve. As long as this does not result in contrast inversion or large color shift or contrast reduction, this will generally not result in significant perceived defects in the observed image for the off-axis viewer 4 .

When the display is operating in the common mode, the set of main image data 6 constituting a single image is input to the control electronics 1 in each frame period, typically in the form of a serial bit stream. The control electronics then outputs a set of signal data voltages to the LC panel 2. Each of the signal voltages is directed to the corresponding pixel electrode by an active matrix array of the LC panel, and the resulting collective electro-optic response of the pixels in the LC layer produces an image. The on-axis viewer 3 and the off-axis viewer 4 then perceive substantially the same image, and the display can be said to operate in a wide viewing mode. This situation is illustrated in Figure 2 and can be said to be a standard method of operation for LCDs.

In the multi-view display device of the type described in GB Patent Application No. 0804022.2 and schematically illustrated in FIG. 3, in the private mode, two image data sets are input to each frame period to Control electronics 1: main image data 7 constituting the main image, and sub-picture data 8 constituting the sub-image 8. The control electronics then outputs a set of signal data voltages, one for each pixel in the previous LC panel. However, the control electronics (display controller) now utilizes an extended lookup table (LUT), and the output signal data voltage of each pixel in the LC panel constituting the combined image depends on both the main image 7 and the sub image 8 The data value of the corresponding pixel (in terms of the spatial position in the image). The output data voltage for each pixel may also depend on a third parameter determined by the spatial location of the pixels within the display.

The output voltage from the control electronics 1 then causes the LC panel 2 to display a combined image that is the primary image when viewed by the primary viewer 3. However, due to the different gamma curve characteristics of the LC panel for the off-axis viewer 4, these off-axis observers most notice the sub-image, which masks and/or demotes the main image, thereby protecting The main image information of the viewer in the restricted cone 9 at the angle of the center of the display. This situation is illustrated in Figure 3. Additional details can be found by reference to GB Patent Application No. 0804022.2.

The image data compression and combination processing procedures and parameters described in the multi-view processing program of GB Patent Application No. 0804022.2 can produce small color artifacts that are apparent to the on-axis viewer 3. This is especially the case for areas of the input main image that consist of a diagonal of a single pixel width. This is due to the fact that this process often results in an output image in which alternating color sub-pixels are set to black, so that the black diagonal of a single pixel width superimposed on this pattern can be retained on either side only A pixel line of one or two color sub-pixels, the sub-pixels being along the length of the pixel line. In this case, the colored lines become visible to the eye.

This problem is illustrated in Figures 4 and 5. 4 shows the results of a multiview image processing program that operates on illuminance values from a 2x2 pixel group of uniform primary and secondary images in a typical RGB stripe display. It can be seen from this that the multiview processor outputs image data that is changed according to the sub-pixel level but maintains the overall illumination and color balance of the 2x2 group. The set of output pixels will then appear to be equal to the input main image to an observer located at a sufficient distance to see only the averaged pixel illumination of the group.

Figure 5 illustrates the same multi-view processing procedure applied to a 2x2 pixel group from an input sub-image, the input sub-images being uniform, but the input main image area consisting of dark diagonal lines. It can be seen that the multi-view processing program results in output image data that differs significantly in both overall illumination and color balance. This difference will be seen by the on-axis observer and will result in a strong green artifact in this example.

The appearance of the image artifact may vary depending on the position of the diagonal relative to the pattern of the light and dark sub-pixels. For example, for example, a dark diagonal in the input main image of the slope opposite the slope shown in Figure 5 would result in an overall magenta appearance in the subset of output pixels. If the multi-view image processing program adds illuminance to all of the color sub-pixels in some of the composite white pixels, and correspondingly subtracts the illuminance from all color components of the adjacent composite white pixels (also as described in GB Patent Application No. 0804022.2) ), instead of patterning based on the sub-pixel level, the diagonal can disappear completely.

GB Patent Application No. 0804022.2 describes the addition of pre-processing steps 10, 11 for the input main image data set and the input sub-image data set in the multi-view image combination processing program (consisting of the image blur filter) to prevent occurrence. These image artifacts. This modification of the display processing procedure is illustrated in Figure 6 of the accompanying drawings (corresponding to Figure 17 of GB Patent Application No. 0804022.2).

In a preferred embodiment of the present invention, an improved image preprocessing method is provided that uses minimal changes to the input image and optimal use of computing resources (eg, by any of the following) One or more to perform processing: fewer gates; less memory for buffering data; less stringent timing; lower latency; and lower power consumption) to be performed by multiview image processing The manipulation of the image data and the knowledge of the image artifacts produced thereby, thereby changing the input image data set in a manner that prevents artifacts from appearing in the output image

In a preferred method, when the input image pixel data is received by the additional image processing device 10 or 11, the input image pixel data is buffered by a plurality of pixels, thereby allowing the "view" or "core" in the image. A subset of adjacent pixel data values is sampled together. Each such window or core is considered to contain a subset of pixel data, wherein each subset contains the same number of pixel groups and each pixel group contains at least one pixel. For example, each pixel group can include a single pixel, or can be a composite color pixel group that includes red, green, and blue pixels. For the sake of brevity, a group of pixels is sometimes referred to simply as a pixel herein.

A three-step handler is then applied to each subset: the first (classification) step is used to represent the pixels within the subset as "high" or "low" values. This sorting step can also be considered as a step of assigning pixels to one of a predetermined set of levels based on brightness (in this case, there are two such levels: "high" and "low"). In the case where the pixel value is associated with a single monochrome or black and white pixel, the pixel values of adjacent pixels are processed. However, in the case of a composite color pixel group (each of which includes red, green, and blue pixels), the process is applied sequentially to each of the color component pixel values of the adjacent composite pixel group.

In the second (comparison or condition detection) step, the resulting high/low pattern is compared to a known pattern set.

In the event that a match is found, a third (operational) step is performed on at least one of the pixels in the subset, and the sampling window is then moved within the image to sample a new set of pixels. In the absence of a match, the pixel value remains unchanged and the sampling window continues to advance (if no match is found, the filter can be applied at least conceptually, the filter simply takes the existing image pixel data and is unchanged In case of it, use it as a new pixel data). Once the window has scanned the entire main input image and performs any necessary changes to the image data values, the adjusted image data is output from the pre-processing circuit to the multi-view circuit.

In an exemplary implementation, the sampling window or "core" is a 4x1 horizontal pixel block. The categorization step involves averaging the data values of all pixels, and defining those pixels in the core having data values above the average value as having a high value (denoted as "1") and having a data value lower than the average value Their pixels are defined as having a low value or "0". This process is illustrated in Figure 7 for an example pixel value column.

Next, the 4x1 pattern of 0 and 1 values is compared with the known pattern set, and the operation corresponding to the current pattern is performed on the second pixel (P _x ) from the left side of the core. Then, the 4x1 window is shifted to the right by one pixel in the image, and the high/low classification, pattern comparison and operation of the second pixel are repeated. This handler continues until the full width of all the columns in the core scanned image. The modified data values from the previous steps can be used in subsequent steps, or the raw data values can be used. For example, in the first phase, it is possible to modify the pixel 2 of the 4x1 window (the second pixel from the left), and the pixel then becomes the pixel 1 in the 4x1 window in the second phase. In the first implementation, the new value of pixel 1 will be used in the calculation in the second phase, while in the second implementation, the original value of pixel 1 will be used. In a video display, the process is performed on-the-fly to filter each frame of the video sequence within the frame time and output the corrected image to the display.

As mentioned previously, in the case where each pixel group is composed of individual color component pixels, the processing procedure is independently repeated for each color component. For example, the description of FIG. 7 can be considered to represent individual values of only red components of four adjacent pixels in a 4x1 window, or a green value, or a blue value. In another embodiment, it will be possible to have the window consider and process values from more than one color component together so that more complex color patterns can be considered.

In an exemplary embodiment, among the 16 possible core patterns, 10 core patterns have been found to require operations on the second pixel from the left to correct image features that would otherwise cause artifacts in the multi-view image. These patterns and corresponding operations are shown in FIG. It should be noted that the pixel value used to calculate P _x ' (the new value of the second pixel from the left in the core) is the original data value of the pixel P _x and its neighbors P _x-1 and P _x+1 instead of A value of 0 or 1 used to represent the high/low value in the first step of the processing procedure, and a new corrected pixel value caused by the previous step if the core is in a different position. It should also be noted from the table in Figure 8, that an embodiment of the present invention is capable of distinguishing between each pattern and its reverse pattern; for example, pattern 1 can be distinguished from the reverse pattern of pattern 1 (which is pattern 2), and the pattern 3 can be distinguished from the reverse pattern of pattern 3, which is pattern 4. Different filters or operations are applied to each pattern and its corresponding respective reverse pattern. One embodiment of the present invention can accomplish this by viewing the pattern in the pixel data rather than the pattern in the absolute difference derived from the pixel data (this is the condition in GB-A-2445982). However, it is not required to distinguish between each pattern and its reverse pattern; it may be the case that the same filter or operation is applied to a particular pattern and its inverse pattern, so in this case it will not be necessary to do between the two Any distinction. It may also be the case that although it is required to distinguish between a particular pattern and its reverse pattern (eg, distinguishing 1101 and 0010), it is possible to apply only a filter to one of the two (eg, 1101), where no filtering is applied. The device is applied to the other of the two (eg, 0010); in this case, only the presence of one of the patterns is required to be detected, rather than the presence of both. In this context, the inverse pattern of pixel data can be considered to be combined to produce a pattern of substantially uniform images, where the light is the opposite of the dark and vice versa.

The parameters (BN, WN, BL, WL, BE, WE, WC, and BC) used to calculate P _x ' can be assigned any value between 0 and 1, and these parameters determine when performing a particular operation. _{The amount by which} the data value of _x is shifted toward the data value of its neighbor. In order to prevent any image artifacts from appearing in the multiview image while minimizing the amount of change applied to the image, these parameters need to be accurately set according to the optical characteristics of the display (brightness, contrast, gamma curve, etc.). In order to optimize the efficiency of the filter handler, it is necessary to make these parameters an integer fraction of 2 ⁿ (ie, m/16 or m/32, where n and m are positive integers) in order to allow the division step to be simple. Bit shift.

An advantage of this embodiment over prior art image blur filters is that the multi-condition detection step allows for determining that a pixel to be operated is part of a line of dark or bright single pixel widths, a single pixel adjacent to a dark or bright line. The line of the width line, the darkness of the two-pixel width or the left edge of the bright line, or the dark-lighted checkerboard pattern of the two-pixel pitch. These are the primary image features that result in artifacts caused by the multi-view image processing program, and each requires different operations to be performed in each condition to optimally correct artifacts that would otherwise occur. In particular, white features can be processed in a different manner than black features, and pixels adjacent to the thin features can undergo different amounts of modification to the pixels that make up the feature itself. This allows the same filter to enhance the appearance of both white and black lines (the ability to significantly increase the perceived quality of the image after a multi-view process). Using different operations for each pattern matching condition (each with individual custom parameters) allows this to be achieved.

It should be noted that although the process is described above as having the core travel left and right across the image, it can be otherwise or vertically moved without changing the overall performance of the process, where Each step makes a corresponding change in the pixel whose data value is changed.

It should also be noted that although the classification steps as described above have been shown to give valid results, there may be specificities for assigning input image pixel data values to the 0/1 state for use with equivalent or more effective There are many other possible ways to compare situation patterns. A simple threshold data value can be applied that classifies the pixel as low when the data value of the pixel is below the threshold and classifies the pixel as high when the data value of the pixel is above the threshold. A threshold having a range can be applied, wherein a pixel having a data value within the range does not have a high or low state assigned thereto. Pixels within a subset may be assigned a high or low value based on their relative data values compared to their neighboring pixels. This method can also be subjected to "spread" in which the pixel is not classified as high or low if the difference between the pixel and its neighboring pixels is less than a certain amount. Such methods, combinations of such methods, and other methods of performing the same classification function are also within the scope of the invention.

It is even possible to dispense with the independent classification step, for example, so that the condition detection step will directly manipulate the image data; the condition detection step will determine which (if any) changes are needed based on the image data itself (eg, via the image with the input as input) Predetermined condition detection algorithm for data (rather than preprocessed classification values). In this case, the first step can be considered to be combined with the second step, wherein the classification and condition detection are performed as a step; basically, this step classifies the pixel data into one of several categories, each category Has an associated filter. It is also possible to combine all three steps into one step; it is necessary that the new pixel data is derived in some way based on the brightness pattern of the subset of pixels, regardless of the steps and steps used to achieve this. situation. Another possibility is that the first step calculates one or more "feature indicators" from the subset of pixels, the feature indicators reflecting one or more corresponding individual metrics associated with a particular known image feature; In this case, one such feature indicator may reflect the extent to which the appearance of the pixel is "checkerboard", where a high value reflects the height of the checkerboard appearance and a low value reflects that the checkerboard pattern does not have much appearance. These "feature indicators" will then be used to determine in the second step which image features (or which features are most dominant) and which filters are best suited for use in the third step.

It will be apparent to the skilled reader that the image processing method comprising the embodiments described above can be implemented with software executing on a processing unit that provides information to a display device. The processing may also be implemented in a special application integrated circuit (ASIC), field programmable gate array (FPGA) (incorporated into existing display control electronics), or other electronic circuits. One advantage of the low computational resource implementation of this embodiment is that the desired circuitry, along with the pixel transistor device, can be placed on the glass substrate of the display, thereby saving cost and allowing the circuit to be easily adapted to each display in which the circuitry is implemented.

An example of a circuit design to perform the required processing is given in FIG. In Figure 9, it can be seen that the processing procedure described above is performed in such a manner that the pixel data values of the image corresponding to the current position of the core are stored in the buffer memory. This value is input to the high/low detector and average calculator for the first (classification) step, and is also input to the parameter multiplication processing program for the third (operation) step. The output of the high/low detector is used in the second (status detection) step by the pattern matching selector to select which of the available parameters input from the register block will be applied to the buffer. The pixel value is used to calculate the correct output value in the third (operational) step. The available parameters are shown in the diagram as WN0/1, BN0/1, WL0/1, BL0/1, WE0/1, BE0/1, WC0/1, and BC0/1, and the 0/1 suffix depends on whether it is for application P _x (the pixel of the current is changed), or a parameter of one's immediate neighbors (e.g., FIG. 8 for the values given, WN1 equal to 1-WN0).

It should be noted that although FIG. 9 depicts the input image data as having a bit depth of 6 bits per color channel (ie, 18 bits per pixel), the figure is only an example implementation and may be The functional handler illustrated in this figure applies to images of any bit depth.

The implementation as depicted also allows for the use of a simple finiteness method in the first (classification) step with the possibility of inputting BT and WT values to determine the data level threshold of the high and low states. This value can also be used in conjunction with the core average to define a range above and below the core average, and the pixel data values must be outside the range to be classified as high or low.

It should also be noted that although the processing procedure as described above produces a significant increase in the perceived quality of the image subjected to the multi-view processing procedure and requires very modest computing resources, variations to the processing procedure are also possible, and such variations can Achieve superior performance at the expense of increased computing resources. As discussed previously, the diagonal lines in the input image are features that typically result in the most noticeable image artifacts in the multi-view image. Therefore, there is a need to use a processing method that distinguishes between diagonal and horizontal and vertical lines, and thus applies the required correction while keeping the off-diagonal relatively unchanged to maintain greater overall image clarity. This is not possible for a single-wire core, but if you increase the core area to cover 2 or more columns, and increase the number of specific patterns used to compare and determine which operation will be applied and the number of available operations The process can then be used to preferentially operate the diagonal while maintaining substantially the same as described above and thereby provide superior image quality in the output multi-view image. The shape of the image scanning core utilized may thus be 4x2, 5x2, 3x3 pixels or any shape found to provide optimal performance without departing from the scope of the present invention.

In a second exemplary implementation of this embodiment, a 4x3 pixel window is used to sample each pixel of the main image and its surrounding pixels in a first (classification) step. Again, in the second (status detection) step, the number of possible conditions and the pixel configuration in the window are compared, and in the third (operational) step, the non-linear data value is combined to the "pixel appearance" conversion step (such as "gamma" Power law conversion) to take advantage of the 3x3 pixel based fuzzy mechanism. Again, in the case of a composite color pixel group (each of which contains red, green, and blue pixels), the process is applied sequentially to each of the color component pixel values of the image. In the case where the pixel values are associated with a single monochrome or black and white pixel, the pixel values of eight adjacent pixels are processed. In this example, a 4x3 window is used in the first step instead of, for example, a 3x3 window as used in the third step, since the 4x3 window should provide a more reliable way of distinguishing between white and black text. However, it should be understood that any size window can be used; and for memory purposes, the smaller the size, the better the two.

The first (classification) step in this implementation may be the same as that used for the previously described method, but is extended for use in a 4x3 pixel window, ie, the average pixel data value of the subset is calculated and will have a higher The pixel value of the data value of the average pixel data value is represented as "1" pixel, and the pixel having the data value lower than the average value is represented as "0" pixel.

In the second (status detection) step, the 3x3 pixel core is used to automatically apply a change to the degree of change to the horizontal or vertical line feature to a greater functional meaning of the diagonal feature. There are fewer individual conditions that need to be detected and that make the application specific to the situation. It has been found that only three general conditions require detection and application of a specific 3x3 pixel based filter operation for each condition. These can be described as a thin feature on a dark background, a thin feature on a bright background, and a checkerboard pattern with a single pixel pitch.

A relatively simple method for determining the condition of two fine features involves summing the number of "1" type pixels in a 4x3 window: if the twelve pixel window contains six or more "1" type pixels, then the condition can be Defined as a dark feature on a bright background. If the twelve pixel window contains less than six "1" type pixels, the condition can be defined as a bright feature on a dark background.

The condition of the checkerboard pattern having a single pixel pitch may be respectively by the left three pixels of the four pixels in the center horizontal line of the 4x3 window (having an alternate pixel classification type, for example, 1 0 1 or 0 1 0) and located at the same The determination is made by three pixels above or below the pixel (having an inverse alternating pattern, for example, 0 1 0 or 1 0 1).

In the third (operational) step, the new data value can be written to the second pixel from the left in the center column of the 4x3 pixel subset, based on the original data value of the pixel and the eight pixels surrounding the pixel ( That is, the original data value of the leftmost nine pixels of the twelve pixels in the 4x3 subset used in the second step. The 3x3 pixel operation used can be as described in GB Patent Application No. 0701325.3: In short, the overall measure of the appearance of each pixel is based on its data value in the main image and its eight The weighted sum of the data values of the neighbors is calculated.

A second overall appearance value for the same subset of pixels is then calculated, but wherein the immediately adjacent neighbors are set to zero to approximate the effect of the multi-view processor on the subset of pixels. If necessary, the weighted sums are normalized and set equal to each other by calculating the new value P' _{(x, y) of the} central pixel data value. This equality is illustrated in FIG. The equation giving the new data value P' _{(x, y)} of the central pixel is therefore:

Where ω ₁ , ω ₂ and ω ₃ are the weighting parameters used, and P _{(x, y) is} the data value of the pixel at the respective coordinates in the main image. As in the previous example, once the new value of the center pixel in the 3x3 window has been calculated, this value is written to the output image ready for input to the multiview image processing program, and the entire three are repeated for the next pixel in the input main image. Step through the process until the entire image has been scanned and processed by the window.

This method can be optimized by selecting pixel weighting parameters to automatically apply a change to a greater degree than the degree of change applied to the horizontal and vertical line features to the diagonal features in the image. This reduction requires the number of specific conditions to be considered in the second (condition detection) step, and as previously described, it has been found that by bright features on a dark background, dark features on a white background, and a single pixel pitch In the three conditions of the checkerboard pattern, only three different 3x3 pixel based operations are applied to achieve an improved image appearance.

In three cases, the weighting parameter of the Equation 1 processing procedure can be determined by taking a 2-D Gaussian distribution of the illuminance of the pixel based on the distance from the center of the pixel in the 3x3 subset, and based on this Gaussian distribution. To quantify the effect of surrounding pixels on the center pixel, as described in GB Patent Application No. 0701325.3. For example, taking a 2D Gaussian with a standard deviation of 0.5 pixel width gives weighting parameters ω ₁ =0.0240, ω ₂ =0.1070, and ω ₃ =0.4759.

In the two conditions of a bright feature on a dark background and a dark feature on a white background, these same weighting parameters can be used in the third (operational) step. However, the degree of change and the overall tendency of the handler to enhance the thin (single pixel) dark or thin features or the appearance of both can be achieved by using the gamma power law before processing the resulting value by a 3x3 pixel based operation. The function first converts the main image data value into a illuminance value, and the gamma power law function is (for example):

Where L = illuminance, D = data value (0-255), and γ (gamma) is a power law coefficient.

If the gamma value (conversion parameter) used in this conversion produces a relationship that accurately describes the response of the data value of the response of the data value of the pixel in the display itself, the operation based on 3x3 pixels will remain as on the display. The overall illumination of the dark features and the illuminated features in the observed main image while still producing a smoothing effect on the clear image features to prevent the introduction of color artifacts by the multi-view processing program.

If the gamma value used is lower than the gamma value of the response describing the display's own data value, then the overall trend of the 3x3 pixel based operation described in Equation 1 will encourage the appearance of dark features in the main image. . This is due to the fact that dark pixels with bright neighbors remain dark after being subjected to the processing procedure, but bright pixels with dark neighbors are substantially darkened.

If the gamma value used is higher than the gamma value of the response describing the display value of the display itself, then the overall trend of the 3x3 pixel based operation described in Equation 1 would encourage the appearance of bright features in the main image. . This is due to the fact that bright pixels with dark neighbors remain bright after being subjected to the processing procedure, but dark pixels with bright neighbors are substantially brighter.

It may be the case that if the appearance of both the thin and the thin features is encouraged by the 3x3 pixel based filter processing program, the overall appearance of the main image after the multiview processing program has been applied is improved. In the case where the second (status detection) step indicates that the 3x3 window is composed of an image area containing a bright feature on a dark background, a high gamma value can be used in the filter operation, and if the second (condition detection) The measurement step indicates that the 3x3 window consists of an image area containing a thin dark feature on a bright background, and a low gamma value can be used in the filter operation to achieve the desired effect under both conditions.

After applying a 3×3 pixel based filter processing procedure to the gamma converted pixel data value, the resulting output value can be converted back to its equivalent new data value using the inverse equation of Equation 2, which is ( that is):

Where D' is the new equivalent data value, L' is the new illuminance value after processing by the 3x3-pixel based filter, and γ is the same power law coefficient as the power law coefficient used in Equation 2.

The effect of a 3x3 pixel based filter operation on two input main image regions (including (a) black diagonal and (b) white diagonal) as shown by Equation 1 is shown in FIG. The pre-conversion and post-conversion are based on Equations 2 and 3, where the gamma values are 1 and 2.4, respectively. From this figure, how the selection of the gamma value affects how the bright and dark lines are changed in different ways, and why it may be necessary to select the gamma value used in the calculation based on the result of the aforementioned condition detection step, so that Encourage the appearance of both thin and thin features and dark features in the image.

The result of the second (status detection) step indicating that the single-pixel pitch checkerboard pattern exists in the region of the main image in which the 4x3 window is sampled in the step can be performed by the equation as described by the following equation Assigning an equal weight to the modified center pixel and its sum of four immediately adjacent pixels performs a simpler 3x3 pixel based weighted sum operation in the third (operational) step:

Again, the pixel data value P _(x,y) can be converted by a non-linear method, such as the gamma dependent handler described above, prior to applying the above operations.

This second example thereby provides for the need to increase the computational resources required for the display (as is the memory required to buffer the 4x3 block of the pixel data value (rather than the 4x1 block) and to perform the necessary pixel data changes. At the expense of the processing power of the operation, in a more precise manner than the first example (this is: the amount of unnecessary changes to the input main image is reduced) and the image removal is caused by the multiview processor. Any color artifacts that are not forgotten by me.

In a third exemplary implementation of the first embodiment, a 4x3 pixel window can be utilized again in the first (classification) step and the second (condition detection) step, and the 3x3 pixel window can be used again in the third operation. However, the number of conditions detected in the second step can be increased from three to six throughout the previous example to allow for greatly simplifying the operations performed in the third step, thereby reducing the need for computing resources of the display, while The minimum blurring effect of the main image to maintain comparable performance (in terms of removing color artifacts from the final multiview image). This increase in the number of detected conditions allows the ambiguity performed in all six conditions to be reduced to a standard normalized weighted sum of 3x3 pixel subsets, where all pixel weights used are integer fractions of 32, At any time, it is not necessary to convert the pixel data value to a illuminance value nonlinearly.

In the first (classification) step, the implementation may be the same as the implementation of the previously described example, ie, calculating the average pixel data value of the subset and representing pixels having data values higher than the average pixel data value as A "1" pixel, and a pixel having a data value lower than the average value is represented as a "0" pixel.

In the second (status detection) step, in addition to the same processing procedure as previously described (compared with known patterns for checkerboard conditions and summed with "1" type pixels in other conditions In addition to detecting the three main conditions of the previous example (the checkerboard pattern, the dark features on the bright background, and the thin and light features on the dark background), the 3x3 sub-to-be used in the third (operational) step is also considered. The central pixel of the collection is classified so that the number of specific conditions is doubled from three to six. For example, a checkerboard pattern having a center pixel of type "0" becomes a different condition in which different specific operational steps are applied to the condition of a checkerboard pattern having a center pixel of type "1". An example of six detected conditions is shown in Figure 12 ("?" in the pixel classification indicates that the pixel type at this location in the subset is not critical to the determined condition).

In a third (operational) step, a normalized weighted sum operation is used to calculate a new value for the center pixel in the 3x3 subset, wherein the different weighting parameters for each of the six different conditions are by the second (condition) The detection step is to identify. In each operation, the weighting parameter can be made an integer fraction of 32 and no non-linear transformation of the pixel data values is required, thereby simplifying the calculation. This process is illustrated in Figure 13, where weighting parameters have been found to produce good results in the final multi-view image for each condition. As can be seen from this figure, each of the six operations is a direct normalized weighted sum, except for the condition "white text neighbors" (where the negative weighting parameters are used for the corner pixels). This allows a large degree of change to be imposed on the center pixel in the event that the center pixel is next to the diagonal, which is needed to remove color artifacts from such areas in the final multi-view image.

An example of a circuit design to perform the required processing is given in FIG. In this figure, in addition to the pixel buffers and arithmetic step handlers that are expanded to have 4x3 and 3x3 pixel windows, respectively, it can be seen that the processing procedures described above operate in a manner similar to that depicted in FIG. For the six different conditions detected in the second step, the parameters input from the register block are marked as A to F in the diagram, wherein 0, 1 or 2 suffix is used to indicate that the parameter is in the first The third (operational) step can be applied to one of a central pixel, one of the lateral neighbors, or one of the diagonal neighbors. For example, if we use the situation detailed as shown in FIG. 13 (where the status = the white character adjacent to "F"), then F0=4, F1=9, and F2=-2.

It should be noted that there are image regions that can cause any of the processing procedures described above to generate new pixel values outside of the normal range (eg, 0-255) under such conditions, such new values can be simply The bit is on the edge of the range.

As an alternative to including a negative weighting parameter, if the method is poor for the particular implementation of the processing program, other methods for applying particularly large changes to the input image material under certain conditions may be utilized, such as filtering. In addition to the operator operation, a constant data amount is added to the calculated pixel value or subtracted from the constant data amount, or the average pixel data value of the 4x3 core (as calculated in the first (classification) step of the processing program) can be used. Make the parameters in the third (operational) step to allow the degree of change to be adjusted based on this average.

It should be noted that although FIG. 14 depicts the input image data as having a bit depth of 6 bits per color channel (ie, 18 bits per pixel), the figure is only an example implementation and may be The function handler illustrated in the figure applies to images of any bit depth.

It will be clear from the above examples that the key purpose of the second (condition detection) step is to distinguish between image regions consisting of bright features on a dark background and vice versa, since it has been found that these different regions are required to be in the third ( Different parameters are applied in the operation step to get the best result. The type of LCD used can make this information available for key instances of these conditions, ie if the image contains black text on a white background, information about this condition can be entered into the LCD, for example, The form of the "flag" embedded in the input image data or as part of the display mode selection by the user. In this case, this information can be entered into a condition detection handler to allow bypassing the first classification step and automatically selecting the correct parameters.

In the second embodiment of the present invention, the multiview processing program is executed on the main image data set 7 and the sub-picture data set 8 in the LCD control electronics 1 without any change to the input image. Then, in the additional image processing device 13, the illuminance value of the output multi-view image and the input main image illuminance value are compared after the compression of the first stage of the multi-view processing program 12 to detect any defects caused by the processing program. Image artifacts. Prior to outputting the corrected multi-view image data set to the display, the additional image processing device 13 then performs some sort of correction on the multi-view image to remove the visibility of the artifact. The processing flow for this operation is illustrated in FIG.

In the exemplary method of the second embodiment, when the multi-view image data and the original image data 7 are input to the additional image processing device 12, the 3x3 core is used to sample each of the two images and all eight of them. Immediately next to each other. In the case where the pixel values are associated with a single monochrome or black and white pixel, the pixel values of eight adjacent pixels are processed. In the case of a composite color pixel group (each of which includes red, green, and blue pixels), the processing sequence is sequentially applied to each of the color component pixel values of the image.

The normalized weighted sum "A" of the pixel data values in the core of each image is then calculated in a process similar to the processing procedure illustrated in Figure 10, wherein the weighting is applied based on the position of each pixel in the core. The pixel. This type of configuration is illustrated in Figure 16, which uses three different weighting factors ω ₁ , ω _{2 ,} and ω ₃ . Comparing the normalized weighted "pixel appearance values" of the corresponding pixel groups of each image, and if a difference is found, the center pixel of the core in the multi-view image is changed by the required amount to weight the weight from each image core Equal to the result. If A _mv is a normalized weighted core sum from a multi-view image, and A _ic is a normalized weighted core sum from the compressed input main image, then the new center pixel of the multi-view image is given by Value P _(x,y) ':

Where ω ₃ is the weighting factor of the central pixel of the core.

Once the new center pixel value has been written to the multiview image, the core proceeds one pixel further and the calculation is repeated, with new pixel values from the previous step providing adjacent pixel values in the new core. In this way, the correction of the multi-view image pixel values is considered in subsequent calculations and the image is not overcorrected. Repeat this process until all pixels in the core have scanned all pixels of the image.

In order to minimize changes to the output multi-view image caused by this processing program, while also minimizing image artifacts in the output image, it is not required to make changes to the uniform area of the main image system. For these reasons, it is also desirable that the weighted core and the uniform main image area do not change with steps (due to the patterning imposed on the multi-view image by the multi-view processing). An example of a uniform region of the compressed main image illuminance value and an equivalent region of the output multi-view image resulting from the processing procedure described in GB Patent Application No. 0804022.2 is shown in FIG. As can be seen from this figure, the 3x3 core at step (n) 14 may have a weighted sum that is different from the weighted sum of the cores at step (n+1) 15, regardless of whether the original image is uniform. Therefore, it is necessary to select weighting values for pixels within the core that result in an equal weighted core sum corresponding to the region of the multi-view image of the uniform region in the input main image. Examples of such weighting are ω ₁ =1, ω ₂ = 2, and ω ₃ = 4.

There is thus provided an image processing method for improving display on a multi-view display device by removing color artifacts otherwise caused by the multi-view processing program while minimizing the degree of change imposed on the input image (especially blurring of fine features) The perceived quality of the image on it. A range of embodiments are described that provide a variety of options for balancing the overall perceived quality of the output multi-view image with the computing resources required to implement the process. It should be noted that although the embodiments provided herein include specific examples of such processing procedures for a particular display device, it has been found that such examples produce good results for different resource specifications. The optimal handler for any particular available resource specification will vary from application to application. Accordingly, particular aspects of the embodiments are desired (such as the precise method of pixel classification used in the first step, the number and type of particular pixel pattern conditions detected in the second step, or the third in the first embodiment) The precise calculations performed in the steps to produce image data changes will vary for each implementation, and when specific specifications for a particular application are known, such changes will be apparent to the skilled practitioner and will therefore be Within the scope of the invention.

An embodiment of the present invention provides an image processing filter that is particularly useful for removing color artifacts caused by processing procedures as described in GB-A-2428152 and GB Patent Application No. 0804022.2. GB-A-2445982 and its non-GB counterparts describe a blur filter for use in a dual view display for processing color artifacts. However, an embodiment of the present invention provides a first blurring filter that is specifically designed to counteract color artifacts caused by the software privacy process, with the result that the amount of blur introduced into the main image is minimized.

However, it should be understood that the present invention is not intended to be limited to software privacy and may be applied to correct artifacts introduced by any image processing step that processes received pixel data to produce a multi-view effect to the viewer. For example, the present invention is applicable to the scenario described in GB-A-2445982, wherein the processing steps considered by the technique according to an embodiment of the present invention are as follows: wherein the interleaving represents different individual The image is received by a collection of pixel data to present different images to viewers located in different respective viewing positions.

Two main embodiments are described above: a first embodiment in which the main image is blurred with reference to itself only before the multi-view processing program is executed on the main image, and the multi-view processed image is compared with the input main image and then relies on comparison A second embodiment of correcting images through multi-view processing is corrected.

In the first embodiment, three examples are provided. It involves three steps: classifying input data, pattern matching, and then applying appropriate fuzzy operations. All three variations have a different balance between performance and ease of implementation/resource requirements.

Some differences between the method embodying the present invention and the method described in GB-A-2445982 and its non-GB counterpart are: one-way status detection filter in GB-A-2445982 and its non-GB counterparts That is, the blur depends only on the pixel value of the pixel being processed and the pixel to the right. A filter in accordance with an embodiment of the present invention uses a pixel to the right, to the left, or both to determine a new value.

A filter in accordance with an embodiment of the present invention is capable of processing black lines, white lines, edges, and checkerboards in a completely different manner. The filter in GB-A-2445982 and its non-GB counterparts has only one specific condition.

GB-A-2445982 and its non-GB counterparts use a differential approach to classify pixels in the core to be similar or strongly different from the pixels being processed. This is insufficient for an embodiment of the present invention because adjacent pixels are more dependent on which operation is performed with a higher or lower value than the key pixel, rather than merely whether it is close or different. In this regard, in one embodiment of the invention, new pixel data is derived for pixels in a manner that distinguishes at least one pattern of pixel data from its respective reverse pattern, which is not the case with the technique of GB-A-2445982.

Some differences between the second embodiment described herein and the second method of GB-A-2445982 and its non-GB counterpart are: this second embodiment uses actual output multi-view images for comparing pixel appearance and Instead of the assumed parallax barrier configuration. Pixels are not always set to zero, others doubling their values, and the pattern of hi/lo pixels is reversed with each movement of the core.

The second embodiment writes the correction to the image, steps the core, and then uses the corrected pixel values from the previous step to calculate a new weighted sum. In this way, the correction is "carry", as described in GB-A-2445982 and its non-GB counterparts.

It will be appreciated that the operation of one or more of the above components can be controlled by a program operating on the device or device. The operating program can be stored on a computer readable medium or can be embodied, for example, in a signal such as a downloadable data signal provided by an internet website. The scope of the appended claims should be understood as covering the operating procedures themselves, or as a record on a carrier, or as a signal, or in any other form.

For a fuller understanding of the nature and advantages of the invention, reference should be

Having thus described the invention, it will be apparent that the same aspects may be varied in many ways. Such variations are not to be interpreted as a departure from the spirit and scope of the invention, and all such modifications are apparent to those skilled in the art.

1. . . LCD control electronics

2. . . LC panel

3. . . On-axis viewer

4. . . Off-axis viewer

5. . . Wide viewing area

6. . . Master image data

7. . . Master image data

8. . . Sub-image data

9. . . Restricted cone

10. . . Image processing device

11. . . Image processing device

13. . . Image processing device

1 is a schematic diagram of a standard layout of control electronics for a liquid crystal display;

2 is a description of a data processing flow in a standard single view LCD;

3 is an illustration of a data processing flow in a multi-view LCD of a known type in a private or multi-view mode;

4 is a schematic diagram showing output pixel data from a known multi-view display processing procedure for area operations on uniform primary and secondary input image data;

5 is a schematic diagram showing output pixel data from a known multi-view display processing procedure for operation of a diagonal region of a main input image and a uniform region of a sub-input image;

6 is a schematic diagram of a process of processing image data in a known multi-view display device having additional pre-processing filters added to the primary and secondary input images;

7 is a diagram showing that in the first (classification) step of the first embodiment, the pixels in the 4x1 pixel sampling window of the main input image are classified as "high" (1) or "low" (0) based on their image data values. Illustration of a preferred method;

Figure 8 is a table showing different operations performed in the third step in ten different situations identified in the second step of the first embodiment;

Figure 9 is a diagram showing possible hardware implementations of the first embodiment;

Figure 10 is a diagram for explaining calculations in a third (operational) step in the second example of the first embodiment;

Figure 11 is a diagram illustrating the results of the calculations of the two input data sets (a) and (b) of Figure 10, wherein different "gamma" parameters are used to nonlinearly pre-convert and post-convert data values;

Figure 12 is a diagram illustrating an example of each of the six different conditions identified in the second step of the third example of the first embodiment;

Figure 13 is a diagram showing possible operations performed on the input main image material in the third step in each of the conditions identified in the embodiment of Figure 12;

Figure 14 is a diagram showing a possible hardware implementation of the first embodiment;

15 is a diagram showing a processing flow of image data in the LCD display device of the second embodiment;

Figure 16 is a diagram for explaining calculations used in the second embodiment; and

Figure 17 is a diagram illustrating the potential of the calculation of Figure 16 depending on the position of the core in the output multi-view image regardless of the potential of the input main image system.

(no component symbol description)

Claims (34)

A method for processing image data displayed by a display device in a display device, comprising: receiving image pixel data representing an image; in a first processing step, processing the pixel data to generate a view to a viewer a multi-view effect; and in a second processing step, each of the plurality of subsets of the pixel data, wherein each subset comprises the same number of pixel groups and each pixel group comprises at least one pixel, dependent Deriving new pixel data for at least one of the groups of pixels of the subset of pixels in the subset of pixels of the subset, the derivation of the new pixel data being capable of distinguishing at least One of the patterns and one of its reverse patterns is performed. The method of claim 1, wherein the first processing step is performed in consideration of a property of the display panel to introduce a change in illumination on the axis that tends to be localized to the on-axis viewer via spatial averaging Balanced and therefore will not be perceived by the viewer on the axis; and changes in off-axis illumination, which will not be spatially averaged to the out-of-axis viewer and will therefore be localized to the off-axis viewer aware. A method for processing image data displayed by a display panel in a display device, comprising: receiving image pixel data representing an image; in a first processing step, processing the pixel in consideration of a property of the display panel Data to introduce the following changes: changes in illuminance on the axis, which tend to be locally balanced by one-axis view via spatial averaging and therefore will not be perceived by the viewer on the axis; and changes in off-axis illumination , which is not locally balanced by the spatial averaging to the off-axis viewer and will therefore be perceived by the off-axis viewer; and in a second processing step, for each of the plurality of subsets of the pixel data In one case, each subset includes the same number of pixel groups and each pixel group includes at least one pixel, and the subset is dependent on the pattern of pixel data in the pixel groups of the subset. At least one of the groups of pixels is used to derive new pixel data. The method of claim 2 or 3, wherein in the first processing step, one of the pair of pixel groups, which is perceived by the viewer as having a single illuminance via spatial averaging, is introduced on the axis Any increase in illumination is substantially matched by substantially equivalent illumination reduction for one of the other of the pair of pixels. The method of claim 4, wherein the obtained illuminance of one of the pairs of pixels is close to a maximum illuminance, or the illuminance of the other pixel group of the pair is close to a minimum illuminance. The method of claim 1 or 3, wherein deriving the new pixel data comprises selecting a filter dependent on the pattern of pixel data, and applying the selected filter to at least some of the groups of pixels of the subset A group of pixels to derive the new pixel data. The method of claim 1 or 3, comprising comparing the pattern of pixel data with a plurality of predetermined patterns, and deriving the new pixel data depending on a result of the comparing step. The method of claim 7, wherein deriving the new pixel data comprises selecting a filter dependent on the pattern of pixel data, and applying the selected filter to at least some of the pixel groups of the subset of pixels Grouping to derive the new pixel data, wherein each of the predetermined patterns is associated with a respective respective filter, and the method includes determining a matching pattern in the comparing step, and selecting an associated filter thereof Used to export the new pixel data. The method of claim 7, wherein the comparing step comprises: determining, for each of the at least one pixel group or the at least one pixel group for which the new pixel data is derived, determining that the pixel group is relatively at least one tight A measure of the brightness of a group of neighboring pixels that distinguishes between higher brightness and lower brightness. The method of claim 7, comprising: assigning, to each of the groups of pixels in the subset, the group of pixels to one of a predetermined set of levels based on pixel data of the group of pixels; and In the step, the pattern of the assigned levels and the predetermined patterns are compared. A method of claim 10, comprising calculating a metric based on the pixel data of at least some of the groups of pixels of the subset, and wherein performing the group of pixels dependent on the calculated metric This step is assigned to one of the predetermined levels. A method for processing image data displayed by a display device in a display device, comprising: receiving image pixel data representing an image; in a first processing step, processing the pixel data to generate a view to a viewer a multi-view effect; and in a second processing step, for each of the plurality of subsets of the pixel data, wherein each subset comprises the same number of pixel groups and each pixel group comprises at least one pixel: based on The pixel data of at least some of the pixel groups of the subset of pixels to calculate a metric; for each pixel group in the subset, based on the pixel data of the pixel group and dependent on the calculated Means assigning the group of pixels to one of a predetermined set of levels; comparing the pattern of the assigned levels to the plurality of predetermined patterns; and depending on the result of the comparing step, among the groups of pixels for the subset At least one of them to derive new pixel data. The method of claim 12, wherein the metric is an average. A method of claim 12, comprising assigning the group of pixels to one of the predetermined set of levels based on a comparison between pixel data of the group of pixels and the calculated metric. The method of any of 3 or 12, comprising, for example, converting the pixel data to a illuminance value for use in the deriving step based on a gamma power law function. The method of claim 15, comprising at least one conversion parameter determined using the pattern of the group of pixels dependent on the subset. The method of any of 3 or 12, wherein the groups of pixels of the subset are connected and extend substantially in one dimension. The method of any of 3 or 12, wherein the groups of pixels of the subset comprise a connected two-dimensional configuration. The method of any of 3 or 12, wherein each of the plurality of subsets is spaced apart from another subset by a group of pixels. The method of any of 3 or 12, wherein the step of deriving the new pixel data uses knowledge of the processing performed in the first processing step. The method of any of 3 or 12, wherein the new pixel data is derived in a manner that takes into account a pattern of pixel data representing image features, the image features being due to the processing performed in the first processing step Tend to cause false shadows. The method of any of 3 or 12, wherein the deriving step comprises determining whether each of the at least one pixel group or the at least one pixel group for which the new pixel data is derived forms a pixel data representing an image feature. A portion of a pattern that tends to cause artifacts due to the processing performed in the first processing step. The method of any of 3 or 12, wherein pixel data representing a bright feature on a relatively dark background is processed in the deriving step in a manner different from processing a pattern of pixel data representing a dark feature on a relatively bright background The pattern. The method of any of 3 or 12, wherein the deriving step comprises: deriving whether each of the at least one pixel group or the at least one pixel group for which the new pixel data is derived from the pattern determination of the pixel data a portion of at least one of the image features: a dark line or a bright line having a width of one of the single pixel groups; a group of pixels adjacent to a dark or bright line having a width of one of the single pixel groups; a dark line or a left edge of a bright line of one of two pixel groups; a dark-lighted checkerboard pattern having a pitch of one or two pixel groups; and a pair of diagonal lines. The method of any of 3 or 12, wherein the second processing step is performed prior to the first processing step. The method of any of 3 or 12, wherein the second processing step is performed after the first processing step. The method of any of 3 or 12, wherein the new pixel data is derived for a single pixel group of the group of pixels of the subset. The method of any of 3 or 12, wherein each group of pixels comprises a composite color pixel group of color component pixels, the method being applied sequentially to each of the color component pixels. The method of claim 28, wherein the composite color pixel group comprises red, green, and blue component pixels. The method of any of 3 or 12, wherein the first processing step is performed to interleave a set of received pixel data for different individual images to present different images to viewers located in different respective viewing positions. A device configured to perform the method of any one of claims 1, 3 or 12. A display device comprising a device as claimed in claim 31. A program for controlling a device to perform the method of any one of claims 1, 3 or 12. The program of claim 33 is carried on a carrier medium, wherein the carrier medium is a storage medium or a transmission medium.